writing-beats
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes a markdown file of "raw material" provided by the user or an external source. This untrusted content is used as a "quarry" to populate the article, creating a surface where malicious instructions embedded in the source could influence the agent's decisions or output.
- Ingestion points: Raw material markdown file (referenced in
SKILL.md). - Boundary markers: The instructions lack explicit delimiters or "ignore embedded instructions" warnings when processing the source material.
- Capability inventory: The skill performs file system reads and writes to manage the article generation process (referenced in
SKILL.md). - Sanitization: No sanitization or validation of the input material is defined before it is paraphrased, quoted, or written to the output file.
Audit Metadata